THE AMERICAN MINERALOGISI" \TOL 54, NOVEMBEII-DECEMBER, 1969

COPPER AND COPPER-IRON SULFIDES AS THEINITIAL PRODUCTS OF SUPERGENE OXIDATION,COPIAPo MINING DISTRICT, NORTHERN CHILE

R. H. Snr.rroE, Deportnlent of Geology, Llniuersity College, Londonr

AND

A. H. Crenx, Department of Geological Sciences, Queen,s [Jni-wrsity, Kingston, Ontario.

ABSTRACT

In numerous deposits of the Copiap6 mining district of northern Chile copper and cop-per-iron sulfides persist as early products of post-Miocene oxidation of hlpogene andsupergene sulfides, the latter formed during a preceding enrichrnent episode. The sulfides ofoxidation origin have been subsequently partialll, altered, predominantly to malachite andgoethite. rn the oxidation zones of this area, hlpogene chalcopyrite has been converted tonormal covellite; hyrogene bornite to "anomalous bornite", idaite, chalcopyrite, and nor-mal covellite; supergene djurleite a.nd hlpogene chalcocite to digenite (or anilite), and toblaubleibender and normal covellite; and supergene digenite to blaubleibender and normalcovellite.

Electron probe microanalysis of blaubleibender covellite revealed a wide compositionalrange, probabll' extending from stoichiometric cuS to at least cur rS, the composition ofblaubleibender covellite s)'nthesized by Frenzel (1961). rdaite is shorvn to have a composi-tion closely approaching cu3leSa, and is consiclered to be quite distinct from syntheticCu; s,Ire,So r-. "Anomalous bornite', in these ores possesses at least 2 r'eight percent lesscopper and 2 percent more sulfur than optically-normal, hypogene bornite. These anal],ti-cal data necessitate the revision of present concepts of the low temperature phase equilibriain this region of the s]'stem Cu-Fe-S.

fN'rnooucrroN

This paper summarizes some of the results of an investigation into themineralogy and controls of the supergene alteration of copper depositsin the Copiap6 mining district of northern Chile (Sil l i toe, 1969a), inwhich the relations between the tectonic and geomorphic history of theregion and the supergene processes have been emphasized. In the morethan 50 years which have elapsed since the recognition and characteriza-tion of the phenomena of supergene oxidation and enrichment, this eco-nomically significant topic has received remarkably l itt le attention.Despit.e the considerable contribution made by supergene alteration tothe mineabil ity of low-grade copper deposits, our widening knowledge ofthe diversity and phase relations of copper minerals stable at low tem-peratures has stimulated few comprehensive studies of supergene as-semblages in individual mineralized areas. For this reason, we feel itworthwhile to present a description of one type of supergene sulfide as-semblage of widespread occurrence, the environment of formation

t Present address: Institu lo de Investigaciones Geol6gicas, Casilla 10465, Santiago, Chile.

1684

COPPER SULFIDES 1685

which has been adequatel-v defined by geologic and geomorphic studl-.

Two distinct "massive chalcocite"l enrichment zones have been dis-

tinguished in the Jurassic to Eocene magmatogene copper deposits of

the Copiap6 region, which include veins, breccia pipes, stockworks,

porphyry copper deposits, and strata-bound mantos' Sulfide enrichment

occurred in the interval between the Lower Eocene and the Upper

Il iocene (Clark, Mortimer, and Sil l i toe, 1968; Sil l i toe, Mortimer, and

Clark, 1968), beneath two erosion surfaces of regional extent, the Sum-

mit and Intermediate Surfaces. I lowever, the 1-oungest and most ex-

tensive major planation surface in northern Chile, the Atacama Pedi-

plain, which developed in response to a third episode of Andean uplift,

everywhere truncated pre-existing supergene alteration profi les. This

phase of pediplanation has been shown to have terminated between 12'6

and 11.5 (+0.5) m.y. B.P. (Clark et a l ' . ,1967). Since the Upper Miocene,

further uplift has resulted in canyon incision, and a fall in the water

table, giving rise to widespread oxidation, and concomitant, quantita-

tively minor deposition of "sooty chalcocite" immediately below the

water table.The sulfide assemblages making up the two zones of "massive chalco-

cite" have been found to comprise dominant djurleite, with minor chal-

cocrte (sensu stricto) and "digenite". The status of this last mineral has

recently been thrown into considerable doubt by the recognition of a

further, distinct phase in the system Cu-S, orthorhombic Cu1 75S, named

anil ite by Morimoto, Koto and Shimazaki (1969). This mineral has

been confirmed as a constituent of "sooty chalcocite" at one mine in the

Copiap6 area (Clark and Sil l i toe, 1969), and probably occurs in minor

amounts throughout the district. Anil ite is converted to a digenite-t1'pe

solid solution bv even gentle grinding at room temperature and, pre-

sumably, by polishing. In the present discussion, the term "digenite" is

reserved for a mineral which is isotropic and bluish in color in incident

iight, and which has a composition intermediate between those of djur-

Ieite and covell ite. It ma1'well be that much of the digenite formed by'

supergene oxidation in the southern Atacama Desert is anil ite. Rem-

nants of hypogene chalcopyrite and pyrite, and of subordinate bornite,

sphalerite, galena, tennantite-tetrahedrite, and enargite persist as cores

in the "massive chalcocite", and become increasingly common with

depth within individual enrichment zones. "Sooty chalcocite" has a

composition similar to that of the massive enrichment assemblages, with

normal covell ite as an additional constituent.

1 A convenient and widely-used field term for compelent aggregatesof supergene cop-

per sulfides, generally having a steely-gray appearance, and contrasting with the pou'der1',

friable accumulations of "sooty chalcocite".

1686 R. II. SILLITO]' AND A. H. CLAIIK

During the development of the Atacama Pediplain, the existing leachedand oxidized zones were denuded, while in some areas practicall l- all theunderlying enriched ore was eroded. Pedimentation resulted in the wide-spread exposure at surface of t 'massive chalcocite" ore, and of the asso-ciated remanent hypogene sulfides. The ensuing oxidation of these sul-fides, which began in the Upper l l iocene, is continuing at the presenttime.

The sulfide assemblages generated in the init ial stages of the in situoxidation of earlier formed "massive chalcocite," and of hypogene chalco-pyrite and bornite which have previously been partially replaced bysupergene copper sulfides, are lhe subject of this discussion.

Mlrrroos or Srunyspecimens of the sulfides and their oxidation products, mounted in cold-setting resin

and polished using ieadJap techniques, were studied by standard optical methods. euan-titative measurements of the spectral dispersion of the reflectivity of several sulfides weremade on the Reichert microphotometer, described by singh (1965). X-ra1' powder studiesof opaque and transparent minerals *ere carried out on a Guinier-de wolfi multiplefocusing camera) using cu Ka radiation. A CAMECA Microsonde Mark r electron probemicroanalyser was used to obtain both qualitative and quantitative compositional data oncarbon-sputtered polished specimens. corrections for absorption, "atomic number factor",and characteristic fluorescence were calculated according to the methods of Duncomb andShields (1966), Duncomb and Reed (1967)1, and Reed (1965), respectively, employing aFortran computer program written by Mr. I{.'G. E. Palfree, of University College, London

OxroetroNl oF D-JURLETTE AND DrGENr.fE

The earliest recognizable stage in the oxidation of djr-rrleite is the for-mation of digenite, "blaubleibender covell ite", and normal covell ite,individually or in association. Both types of covell ite may also have beenderived from the oxidation of digenite of enrichment origin. ,,Massive

chalcocite" has locally been altered to these copper sulfides alone, butthey also form as an intermediate stage in the oxidation of djurleite ordigenite to malachite or, rarely, atacamite, or to the association ofmalachite and goethite known locally as almagrad,o. In many specimens,however, almagrado formation has apparently not involved the priordeposition of copper sulfides. "Massive chalcocite',, r immed and veinedby cuprite, native copper or tenorite, occurs in the Copiap6 region, butcovell ite of oxidation origin is never present in such ores. However, asingle specimen of djurleite rimmed by delafossite exhibits normal covel-l i te as an intermediate oxidation product (Sil l i toe, 1969b).

The <lptical properties of normal covell ite are distinct from those of

I Duncomb, P. and S. J. B. Reed (1967) The calculation of stopping power and back-scatter effects in electron probe microana\ysis. Tube Inrestments Research Lab., Cambrid,ge,Eng., T ech. Rept.,22l.24 p. (freely available from company or authors) .

coPPER SULFIDES 1687

the "blue-remaining" variety, which was first iully characterized by

Frenzel (1959). Blaubleibender covell ite is distinguished b-v the absenceof a purple-red (R) coloration in plane-polarized incident l ight in oil

immersion; it shows bireflectance from light to dark blue in air, but this

effect is only slightly intensified under oil. The anisotropism is also weaker,exhibit ing orange-brown hues in the place of the distinctive fiery oranges

and red-browns of normal covell ite. The reflectivity and polishing hard-ness of the normal and blaubleibender covell ites are of the same order.

Clear evidence of the existence of varieties of covellite having opticalproperties intermediate between those of the normal and blaubleibendermodifications has been obtained from optical study of specimens fromthe Copiap6 district, but the areas covered by these are too small topermit more detailed examination. The X-ray powder patterns of the

two main types are similar to those listed by Frenzel (1959). The in-distinct, diffuse lines show few appreciable differences in d-spacings,although as, pointed out by Moh (1964), blaubleibender covell ite pat-

terns Iack several weak lines which characterise the normal form.Digenite of undoubted oxidation origin was only rarely observed in the

specimens examined, but, especially in ores from the Quebrada Puquiosdistrict, residual grains of djurleite (10-20 pm in diameter, and enclosedby goethite and minor malachite) are seen to have been altered to dige-nite, in turn oxidized to both blaubleibender and normal covell ite. It was

diff icult in some specimens to ascertain whether digenite had formed

contemporaneously with the associated djurleite during enrichment, or

was a product of subsequent oxidation. The consistent occurrence in the

Puquios ores of digenite only in association with djurleite which has

undergone oxidation, is considered to reflect its role as the first stage in

the oxidation of djurleite, preceding the appearance of covell ite.Blaubleibender and normal covellite are more widespread than dige-

nite, and were observed in practically all the oxidized ores examined in

the present study. The covellite generally takes the form of flame-shapedlamellae or minute patches, some 5 to 10 pm in diameter, in djurleite and

digenite. No covell ite has been distinguished in specimens of oxidized

supergene chalcocite (sensu stricto), although hypogene chalcocite, asso-

ciated with bornite in disseminated deposits in volcanics, has locally been

replaced by normal covell ite. The covell ite flames are generally concen-

trated in the vicinity of cross-cutting veinlets of gangue minerals, frac-

tures, or the boundaries of "massive chalcocite" aggregates, but are alsopresent within apparently unfractured copper sulfides. Crystallographiccontrol of covell ite formation is evident in some samples. In specimens

which exhibit remanent cores of hypogene sulfides, a zor\e of unaltered

"massive chalcocite" may be present between each sulfide core and sur-

rounding covell ite-bearing areas.

1688 R. H, SILLITOE AND A H, CLARR

Where both digenite and djurleite are present in a single polishedsection, the former has been preferentiall-t ' replaced by normal andblaubieibender covell ite. Blaubleibender covell ite is, however, more con-sistently associated with digenite than is the normal varietl . This rela-tionship has been recorded b1,' several previous workers (e.g. Vokes,1957;Frenzel, 1959; Ramdohr, 1960; Grafenauer, 1963). More extensivedjurieite areas, several mm in diameter, hoy be wholly replaced bymosaics of both covell ite modifications, and exhibit irregular cracks, dueto the volume-reduction which accompanies the replacement (Figure 1).The microscopic relations indicate that, where both normal- and blau-bleibender covell ite occur as the products of djurleite oxidation, thelatter formed earlier. Thus, remanent cores of djurleite in replacivegoethite and maiachite are altered at their centers to flames of blaublei-bender covell ite, while they are replaced marginally by the normal vari-ety. Residual grains of normal covell ite are also retained in the adjacentgoethite malachite intergrowths. Normal corell i te formed by the oxi-dation of "massive chalcocite" occurs commonly as grains in goethite,malachite, and areas of gangue, whereas the blaubleibender form hasonly rarely been recognized in such contexts.

Digenite and covell ite, intermediate products in the transformation of"massive chalcocite" to malachite or to aggregates of malachite andgoethite, are found throughout the vertical extent of oxidized zones, butare especially concentrated in their near-surface sections. "Massive chal-

Frc. 1. Djurleite (light gray), partly altered to blaubleibender covellite (dark gray)Note the cracks due to volume reduction. 25-m level, Mina Santos, Punta del Cobre dis-trict. Plane polarized incident light; oil immersron.

COPPIlR SULFIDI':,5

cocite" altered to covell ite alone occurs in manl' zones o{ preferentialdownward water percolation, close to the lower l imit of oxidizing influ-ences. Blaubleibender and normal covell ite are consistently absent in"massive chalcocite" ore which has not undergone a subsequent phaseof oxidation.

In hand specimen, it is generally impossible to distinguish covell itewhich has formed as an intermediate stage between "massive chalcocite"and malachite and goethite. Competent enriched ores which have beenaltered to covell ite superficially resemble "sooty chalcocite". However,such oxidation assemblages are characterized by a gray-black color, con-trasting with the black or bluish-black hue that typifies powdery sulfidesformed by enrichment, and by a cavernous, slaggv texture, which re-flects the volume decrease during replacement.

Much of the normal, and all of the blaubleibender covell ite whichoccurs in the Copiap6 region has been formed through oxidati,on. Aminor proportion of the normal covell ite is a constituent of true "sootychalcocite", precipitated by the enrichment of chalcopyrite and pyrite.No covell ite of enrichment origin has been recognized in "massive chal-cocite" ores from the district. The origin of some or all normal covell iteby the oxidation of "chalcocite" has been noted at Chuquicamata, An-tofagasta Province, northern Chile (Bandy, 1938), at the CampbellMine, Bisbee, Arizona (Schwartz and Park, 1932),in the Globe-Miamidistrict of Arizona (Peterson, 1954, 1962), at Castle Dome, Lrizona(Peterson et aI., I95l), and at Kennecott, Alaska (Bateman and Mc-Laughlin, 1920). However, the formation of digenite in the course ofsupergene oxidation has apparentll. not been described from naturalassemblages.

The composition of blauble'ibender coaellite

Quantitative electron microprobe analyses have been made of severalspecimens of blaubleibender covell ite from Mina Santos, a small minein the productive Punta del Cobre group, using optically normal covell itefrom Mina FIor del Llano as a standard; both covell ites occur as earl1,oxidation products of "massive chalcocite" (djurleite) rimmed b.v mala-chite. The analytical results are presented in Table 1. No anall 'ses weremade of the small areas of covell ite having optical properties interme-diate between those of the normal and blaubleibender varieties. Con-siderable diff icult-v was encountered in finding grains of blaubleibendercovell ite of sufficient purit l ' for anallsis, but it is considered that thesedata give a good indication of the composition of this phase.

Tt is evident that the blaubleibender modification of covell ite in thisregion contains between 2 and, 7 rveight pereent more copper than does

1689

1690 R I SILLITOE AND A, II . CLARK

Tesln 1. ElrcrnoN Pnoln MrcnolNALysES oF Br,Aunt,nrsrNntn Covor-r-trn,MrNa Se,Nros, PuNre onr Cornr, Coerae<l, Cnrr,n

Sample Wt. % Cu Wt. /6 Fe w t % s

I2J

A

.)6

7 0 . 269. 87 0 . 168. 76 9 . 57 3 . 3

1 i

o . /

7 4n.d .

28.725.62 7 . 428.9

100. 3102.19 8 99 7 . 6

CuS 66.48Synthetic blau-

bleibendercovellite(Moh, 1964) 67.5-68.5

Synthetic blau-bleibender

covellite(Frenzel , 1961) 73.25 0 1 0

33 .52

31 .5 -32 5

26.45

100.00

99.89 .

" Includes 0.09 wt. o/n I{rO.

normal covell ite, in addition to a small amount of iron. The observediron content (determined using associated chalcopyrite as a standard)may have been contributed in part by minute inclusions of goethite, andwas almost certainl). due to this cause in Sample 2 (Table 1), but someat least of the blaubleibender covell ite is ferroan. We have assumed inour analytical work that the normal covell ite standard is stoichiometricCuS, as found by Kullerud (1965), and smail errors could have beenintroduced by any departure from stoichiometr), ' .

Frenzel (1959) originally considered blaubleibender covell ite to be1.5 to 2 percent richer in copper than normal covell ite, but the subsequent(1961) analysis of synthetic material showed it to be as copper-rich asCu1.aS, essentiaily the composition of our analysed Sample 6 (Table 1)Moh (1964) delimited, by hydrothermal s)'nthesis under oxygen-free.conditions, an apparent stabil ity f ield for this mineral, extending fromca. 67 .5 to 68.5 copper at 50oC and 135oC, or one to two percent morecopper-rich than CuS.

The microprobe analyses presented here demonstrate that the blau-bleibender covell ite from Mina Santos achieves compositions more thanfour percent richer in copper than that synthesized b1'Moh (1964). It isevident that, under natural conditions, blaubleibender covell ite mayhave an extremelv variable composition, probably ranging from stoichio-metric CuS to at least Cur.nS;if the optical properties of this phase are

COPPER SUL],'IDI;S

directly influenced by its Cu:S ratio, such wide compositional variationsappear to be represented in very restricted areas. The status of blue-remaining covell ite as a phase in the svstem Cu-S remains unclear, butit is considered that the field of stabil ity put forward b1' Moh does notadequatell ' reflect the compositional relations of the natural mineral.The blaubleibender phase (Cu1.aS) s1'nthesized by Frenzel (1961) may,on the other hand, indicate the l imit of tolerance for copper of the covel-l i te structure at low temperatures. Whereas Moh (1964) has concludedthat, in view of its higher symmetry' relative to that of normal covell iteand its apparentll ' distinct composition, the term "blaubleibender covel-l i te" is misleading, rve feel that this convenient description should beretained unti l the crystal structure and stabii ity f ield of this phase areclarified.

Oxrrarrow oF CHALCoPYRTTD

Residual grains of chalcopl-rite, generally marginally replaced bydjurleite, have been oxidized in sitw in many ores to aggregates of mala-chite and goethite, which in hand-specimen have a deep-brown color anda speckled appearance, owing to the presence of areas of a blackishpitch-like goethite. In polished section, small spikes and patches of nor-mal covell ite may be observed as intermediate products of this oxida-tion, rimming the chalcopyrite, and penetrating inwards along (111)cleavage directions. The chalcopyrite residua are eventually pseudomor-phosed by goethite, which may retain lamellae of normal covell ite.Covell ite occurring in the surrounding goethite may have been derivedfrom the oxidation of either the chalcopl'r ite or djurleite. This normalcovell ite of oxidation origin is readily distinguished from that formed(as "soot1' chalcocite") in the enrichment of chalcopyrite, which is un-accompanied by other contemporaneous minerals, and criss-crosses chal-copyrite grains in a network of cleavage-controlled laths and veinlets.

Oxrl.qrroN or BonNrrr

The ultimate products of the oxidation of bornite, malachite andgoethite, are similar to those resulting from the oxidation of djurleiteand chaicopyrite. However, the intermediate stages in the alteration ofbornite involve the formation of a chalcopyrite-l ike phase, idaite, and abornite-l ike phase, here referred to as "anomalous bornite", in additionto covellite.

Idaite was first named and comprehensively described by Frenzel(1958, 1959, and 1960), although an apparently similar phase in thesystem Cu-Fe-S had earlier been synthesized by Merwin and Lombard(1937) and by Roseboom and Kullerud (1958). Frenzel described idaite

1691

t692 R IT. SITIIIOE AND A. H. CLARK

as a supergene sulfide which is formed, characteristically, by the altera-tion of bornite. This mineral has since been recognized, in associationwith bornite, at Yauricocha, Peru (Kobe, 1961), in various deposits inArgentina (Brodtkorb, 1961), from deposits in Japan (Takeuchi andNambu, 1961), at Lit i ja, Yugoslavia (Grafenauer, 1963), at Sommerkahl,Germany (von Gehlen, 1964), at Konnerud, Norway (Krause, 1965),and at Trattenbach, Austria (Tufar, 1967). The optical properties ofidaite from the Copiap6 region are distinctive, and closely comparableto those described b1' Frenzel (1958 and 1959). The bireflectance in airis strong, from reddish-orange to gray-yellow; oil immersion enhancesthese colors. The anisotropism of idaite is enormous, ranging from palegreen to yellow-green in air, while the reflectivity at 589 nm is approxi-matell ' 23 to 32 percent.

The anomalous bornite distinguished in the present study has a moreyellowish-brown coloration than the normal bornite in these and otherdeposits, and some grains are distinctly anisotropic in air, from red-brown to gray-brown. The white-l ight reflectivity appears to be slightlyhigher than that of the normal variet1., an impression confirmed byquantitative refl ectivity measurements. The refl ectivit l ' dispersion curvesof normal and anomalous bornite (Fig. 2) differ significantly, the anom-

4(Jo 440 4n 520 560 600 640 680trnm

Frc 2. Reflectivity dispersion profiles for anomalous (A) and normal (B) bornites fromX{ina Esmeralda, Cerro Blanco district. Each curve represents the mean of four sets ofvalues. (xReflectivity relative to Royal School of Mines carborundum standard (N.P.L.calibrated-see Singh, 1965,)

I

is(l

t5

COPPER SULFIDES

alous phase having a reflectivity lower at short wavelengths and higherat long wavelengths than that of opticallv normal bornite. Apparentll 'similar, opticall l ' unusual varieties of bornite have been observed tooccur in close association with idaite by Frenzel (1959), Brodtkorb(1961), Tufar (1967), and von Gehlen (1964), who alone, before thepresent studv, investigated its composition.

The oxidation of hypogene bornite, earlier rimmed and veined bysupergene djurleite, to yield, successively, anomalous bornite, the chal-copyrite-l ike phase, idaite, and covell ite, has been studied in detail inpolished sections of ores from the Jardinera, Abundancia de Puquios,Manto Esperanza, and Esmeralda mines in the Copiap6 area. Thissequence of mineralogical transf ormations is represented whereverbornite has been exposed to oxidation, but is absent from assemblageswhich have been affected by sulfide enrichment alone. The generalizedalteration process has been diagrammatically represented in Figure 3(A - I ) .

The init ial stages of alteration of the bornite and associated djurleite(Fig. 3A) involve the marginal conversion of djurleite to goethite andmalachite, and the formation of patches and flames of blaubleibender ornormal covell ite in the djurleite where it abuts against the replacivegoethite and malachite (Fig.38 and 3C). Minor normal covell ite maydevelop in the malachite and goethite close to the boundary with thedjurleite at this stage (Fig.3C and Fig. a). The bornite at f irst remainsunaltered, but the oxidizing solutions then gain access via microscopiccracks in the narrower djurleite rims, or on the total replacement of thesurrounding djurleite by the oxidate assemblages.

Where the bornite has not been partiallv replaced by djurleite in apreceding enrichment episode, its oxidation commences immediately.Where a narrow ((5 p-) djurleite rim is preserved, small tufts andpatches of the chalcopyrite-l ike phase, sometimes bordered by normalor blaubleibender covell ite, may form along its contact with the bornite(Fig.3D). The chalcopyrite, which differs in no way from hypogenechalcopyrite in its general optical properties, locally penetrates inwardsalong the cleavage planes of the bornite. In some specimens, the break-down of bornite and djurleite to chalcopyrite and covell ite has led to theformation of a distinct halo of these two sulfides at the margin of bornitegrains, enclosed in remanent "massive chalcocite" (Fig. 3E). Very oc-casionally, sparse Iamellae of idaite accompany this chalcopyrite.

The succeeding stage in the alteration of the bornite involves thegradual accession of the optical characteristics of anomalous bornite, andthe marginal development of bundles of idaite and chalcopyrite lamellaealong cleavage planes in the bornite (Fig. 3F, and Fig. 4 and 5). In some

1693

1694

A-

R. .q. SILLITOL' ,4ND A II . CLARK

F

3O/:m,

Frc. 3. An idealized oxidation sequence for bornite previousl-v partially replaced

by djurleite.

COPPER SULFIDES 1695

Frc. 4. Bornite (medium-gray) enriched by djurleite (white-top left) and oxidized to

goethite (light gray) and malachite (black). A grain of normal bornite (lower middle of

field) has suffered oxidation to anomalous bornite (medium-gra1'), rimmed by lamellar

idaite (whitish) and normal covellite (dark gray). Scattered grains of djurleite and normal

covellite occur in the surrounding malachite. 6-m level, Mina Jardinera. Plane polarized

incident light; oil immersion.

specimens, idaite alone is present at this stage (Fig.5), but in otherschalcopyrite is well represented. The chalcopyrite has a tendency to oc-cur, generally as isolated lamellae, close to the centers of grains of ano-malous bornite, while the idaite is concentrated at their margins. Theidaite lamellae average 23 to 45 pm in length, and, with continued oxida-tion, advance inwards at the expense of the anomalous bornite unti l theaggregate consists predominantly of idaite (Fig. 3G), with or withoutchalcopyrite. The chalcopyrite and idaite are often interleaved withgeothite or normal covell ite (Fig. 5), and, rarely, with blaubleibendercovell ite, giving a striped appearance in polished section. A rim of nor-mal covell ite, with minor goethite, may separate the zone of chalcopyriteand idaite from the surrounding goethite and malachite (Fig. 3F and3G; Fig.5). Veinlets of normal covell ite occasionally cross-cut the idaiteand chalcopyrite lamellae and the outermost parts of anomalous bornitegrains.

The idaite and chalcopyrite are subsequently transformed to goethite(Fig.3I), which is at f irst intergrown with minor normal covell ite (Fig.

3H). In the course of the sequence of mineral replacements describedabove, the volume occupied by sulfides gradually decreases in favor ofeoethite and malachite.

1696 R. I]. SILLITOE AND A. II CLARK

Frc 5. Anomalous bornite (medium-gray) rimnied by lamellar idaite (whitish) andinterleaved normal covellite (dark gray) and malachite (black).'fhe idaite is rimmed bynormal covellite (dark gray) and malachite (black). 6-m level, Mina Jardinera. Planepolar ized incident l ight ; o i l immersion.

Frenzel (1959) originally described idaite as a product of supergeneoxida,tion, but, although later workers have recognized its supergeneorigin, the general textural relations have led to the erroneous conclusionthat this sulfide has formed during enrichment (e.g. Takeuchi and Nam-bu, 1961; Grafenauer, 1963; Krause, 1965). In the assemblages describedby these authors, bornite is rimmed by various associations of idaite,chalcopl'r ite, and covell ite, and in some cases, by an outer zone oI "chal-cocite" or "digenite". This sequence (c/. Fig. 3E) has given rise to theassumption that the chalcopl-rite-idaite-covell ite association is an inter-mediale stage in the formation of the enriched ores. It is, however, clearthat these phases are formed during oxidation. Such an assemblage istvidespread near surface in the Copiap6 region, but is invariably absentin enriched ores which have not undergone subsequent oxidation.

Composit'ions oJ anomalous bornite antl ,idaite

Electron probe microanalyses have been made of associated normai,hypogene bornite, anomalous bornite, and idaite in three specimens fromthe Abundancia de Puquios and Esrneralda (Cerro Blanco) mines. Theanalytical data are presented in Tables 2 and 3, and in Figure 6. Threeanalyses of normal bornite f ield compositions in the immediate environsof stoichiometric CusFeS+ (Table 2), in agreement with the conclusionsof most earlier workers (e.g. Zeis and Merwin, 1955; Brett and Yund,

COPPER SULFIDES

Tr\BLr 2. Et-tcrnoN ProBE MrcronNALysES ol Nonlr,q,l eNo

Awouar,ous BonNrrn, Coprlp6 Drstrrcr

r697

SamplewL ok wt.% wL ak

C u F e STotal Locality

Hypogene bornite

Hypogene bornite

Hlpogene bornite

Cu;FeS+

6 3 . 0 6 1 1 . 8 4 2 5 . 7 9

63.40 11 .62 25 .87

63.79 11 .84 26 55

6 3 . 3 3 1 1 1 2 2 5 . 5 5

99.69 Mina Abundancia de

Puquios

100.89 Mina Esmeralda(surface)

102 18 Mina Esmeralda(10 m level)

100.00

Anomalous bornite

Anomalous bornite

Anomalous bornite

Anomalous bornite

61 .13 12 .06 27 80

61.20 12.63 27 49

6 0 . 8 r l . 4 2 6 . 2

6 r | 1 1 . 5 2 9 . 0

100.99 Milra Esmeralda(surface)

101.32 Mina Esmeralda(10 m level)

98.4' Mina Esmeralda(10 m level)

101 .6' Mina Esmeralda(10 m level)

Anomalous bornite 61 .5(von Gehlen, 1964)

1 1 5 2 6 . 5 99. 5 Sommerkahl(Spessart)

u Average of two discrete areas.

_T.lnm 3. Er-rcrxoN Pxo.er MrcnoaNALysES on Ilelrr, Copr,tp6 Drsrnrcr

SampleWL o/o WL. "k wt. 9'o

Cu Fe , Total Locality

Idaite

Idaite

Idaite

.53 78 15.28 33.26

5 r .71 15 .11 33 .80

52 .87 16 .52 33 .78

102.32 Mina Abundancia de

Puquios

100 62 Mina Esmeralda(surface)

103.17 Mina Esmeralda(10 m level)

CuaF-eSr 50.87 14.9C 34.23Cu;IreSa 56 14 9 87 33 99Cu; rFeSo ; 56 .94 9 . l1 33 . 95" Ida i te " 56 .3 9 .8 33 .7

(F renzel andOttemann, 1967)

100.00100.00100.0099.8 Nukundamu, Fi j i

1698 R. II , ilLLITOIi AND A. N. CLARK

Cu5FeSu Wt lFe

Cu., FeSo., .( synrheric - t50 t) Cu.FeSa

!- Supergene\ lda i tes,

Atacama

Anomalous30 ca

+ sBornites,

Atacama

X-Bornite,60"c

Cu5FeSa - Hypogene Bornite,Atacama

Frc. 6. Compositional relations in part of the system Cu-le-S, illustrating the lack ofcorrespondence of the anomalous bornite and idaite from the Copiap6 region with knownsynthetic phases. The boundaries of the bornite-digenite solid solution at 700.C (Yundand Kullerud, 1966) and the composition of the idaite-like mineral described by Frenzeland Ottemann (1967) are included for comparison.

1964). CelI edge determinations were carried out on four specimens ofoptically normal, hvpogene bornite from the Copiap6 area, by measure-ment of the angular position of the (440) reflection, using a Norelcodiffractometer, Cu Ka radiation, sil icon internal standard, and an oscil la-tion technique. The measured values of 10.950, 10.950, 10.952, and10.952 A (+0.001 A) are essentiallf identical to that of stoichiometricCusFeS4, and are evidence of the compositional homogeneily of normalbornite in these deposits.

The differences in the optical properties of normal and anomalousbornites were considered by Brodtkorb (1961) to be due to the presencein the latter of submicroscopic inclusions of chalcopyrite. However,microprobe line scans across several areas of anomalous bornite haveshown that, at least in the Copiap6 ores, this phase is homogeneous withrespect to copper, iron, and sulfur at a beam resolution of ca. 1.5 pm.Four microanalyses of the anomalous bornite (Tabie 2, and Fig. 6)indicate that this phase contains up to three weight percent more sulfur,and is consistently richer in iron and poorer in copper than the opticallynormal bornite. Analyses standardized against metall ic iron and copper,and pyrite, and against the associated normal bornite yielded essentiall l 'identical results. The four analyses outl ine a compositional f ield lying

COPP]JR SULF'IDES

beyond the boundaries of the bornite-digenite solid solution field at700"C (Figure 6). It is clear, moreover, that, with increasing sulfur con-tent, the anomalous phase takes on a more yellowish coloration in planepolarized l ight, and progressively becomes more distinctly anisotropic.

These observations clearly invite comparison of the anomalous bornitewith the sulfur-rich natural bornites discussed at length by Brett andYund (1964) and by Yund and Kullerud (1966), and with the "X-bornite" phase synthesized by the Iatter authors. In their work on thesystem Cu-Fe-S, Yund and Kullerud describe the ready formation of"X-bornite" at temperatures below ca. 140oC. This phase contains ap-proximately 0.4 atomic percent more sulfur than stoichiometric Cu5FeSl,and a portion of its apparent stabil ity f ield at 60oC is indicated in Figure6 for comparison with the present data. Although stating that "X-bornite" is optically very similar to normal, tetragonal bornite, Yundand Kullerud give no details of its properties; this sulfur-rich phase does,however, yield a distinct X-ra-v powder pattern. Unfortunately, no com-parative X-ray data could be obtained for the anomalous bornites fromthe Copiap6 region because of the small size of the available specimens.

The natural and synthetic sulfur-rich bornites investigated by Brettand Yund (1964) were found to exsoive chalcopyrite, and sometimesdigenite, when heated to temperatures of 75oC and over. In aacuo heatingof a sample of anomalous bornite from Mina Esmeralda similarly yieldeda small number of chalcopyrite lamellae after seven days at 115oC, sug-gesting that this mineral is stable only at lower temperatures. Yund andKullerud (1966) X-rayed a natural "anomalous bornite", and found itto consist of a "mixture" of their "X-bornite" and normal tetragonalCurFeSr. The anomalous bornite which has been recognized at severalIocalit ies in the Copiap6 mining district might, therefore, be equated withsuch a mixture, if i t were not for the marked compositional disparitybetween the Copiap6 phase and "X-bornite" and normal bornite. Themineral under consideration appears to be significantly richer in bothsulfur and iron than any of the sulfur-rich bornites observed by earlierworkers, with the exception of that briefly described by von Gehlen(1964) from the Sommerkahl deposit, Spessart. Von Gehlen obtained, b1'microprobe analysis, a composition almost identical to that of one of theCopiap6 specimens (Table 2). He found that annealing at (apparently)75oC caused the development of small bodies of a chalcopyrite-l ikephase, which, however, had an approximate composition of CusFezSz.

Correlation of the Copiap6 and Sommerkahl anomalous bornites withthe known low temperature phase equil ibria in this region of the systemCu-Fe-S encounters serious diff iculties. Although they established thepresence of the sulfur-rich "X-bornite", Yund and Kullerud (1966)

r699

1700 ]t. II. SILT,ITO]| AND A. II CLARK

determined its stabil ity l iniits onll ' along the CurFeSr-CurFeS6 "join",and the effects of varving the copper:iron atomic ratio were not in-vestigated. The coincidence of the present analyses with that of vonGehlen lead us to propose that these data outl ine part of a true field ofstabil it l- for the anomalous bornite phase at approximately 25oC. Thisphase ma1' or may not be identical to, or continuous with, "X-bornite",and further low temperature s1'nthetic work is required to delimit theirrelations. At the present t ime, however, Yund and Kullerud's assertion(1966, p. 484) that "all anomalous bornites contain the X-bornite phase"

surely remains unfounded.From the three microanalyses of lameilar idaite grains (Table 3 and

Fig.6) , i t is ev ident that th is mineral conta ins, on average, some 3.5weight percent less copper, and six percent more iron than are representedbv the formula, Cu5FeS6, given for idaite by Frenzel (1959). The Co-piap6 sulfide is even more divergent in its composition from the generalformula, Curs,Fe,Sorr l proposed by Yund (1963). A mean (CuiFe) :S

atomic ratio of 1.06 is f ielded b1.the three new analyses (recalculated to100 percent), in close agreement with that of the idaite-l ike phase s-vn-thesized by Yund and Kullerud (1966), but the analv-ses give a meanCu:Fe ratio of 2.97,radically different from that of the synthetic phase.The compositions of the three analy'sed idaites ma1' be represented bythe formulae: Cur. rFeSs.s; Cua.oFeS3.e; and Cuz.sFeSs.o. This minerai ,therefore, has a composition closell approaching stoichiometric CuaFeSq,albeit showing a slight, and perhaps significant, sulfur-deficiency. Minuteinclusions of covell ite and goethite may have contributed minor errorsto the anal-v-ses, but the coincidence of the compositions determined forthe opticall l ' normal, h1'pogene bornites (Table 2) with stoichiometricCusFeSa, encourages confidence in the accurac)' of the new data. It isconcluded that the mineral occurring in these oxidized assemblages iscompositionally quite distinct from the phase s1-nthesized in the systemCu-Fe-S.

These observations are in excellent agreement with those recentlvpresented by L6vy (1967), who concluded, on the basis of f ive microprobeanalvses of reasonably pure specimens from two localit ies, that idaite,occurring in association with bornite and presumabll ' of supergene origin,has a composition of Cu3FeSr. L6v1' demonstrated that natural idaite

f ields spectral reflectivity dispersion curves difiering markedly f romthat of the original "CusFeS6" s1'nthesized b1. X'Ierwin and Lombard(1937). We have confirmed L6v1,' 's observations on the reflectivity ofidaite through measurements on specimens from the Manto Esperanzzrmine (Figure 7). Two areas of poiygranular idaite aggregates 1-ieldeddispersion curves essentially parallel to those found for pure idaite fronr

COPP]'R SULFIDE^S

450 500 550 600 650 700) nnr

Frc. 7. Reflectivity dispersion profiles (dashed lines) for idaite from N{ina N{anto Esper-anza, Copiap6 district, with that for idaite from Bancairoun, France (L6vy, 1967), for com-parBon.

Bancairoun, France (L6vy, 1967), but lving at slightl l ' lower values. Wehave been unable to obtain reliable X-ray powder data for idaite fromthe Copiap6 ores, but L6vv, who encountered similar diff iculties, haspointed out that the powder data given by Frenzel (1959) in fact bearclose comparison with the tetragonal pattern of mawsonite, an inter-mediate member of the idaite (Cu3FeS+)-stannite (Cu2SnFeSa) series.

In contradistinction to the evidence obtained from the microprobeanalysis of natural, supergene idaite, Frenzel and Ottemann (1967) haverecently- demonstrated that a hypogene, idaite-l ike phase occurring inthe Nukundamu deposit, Fij i , has a composition (Table 3; Fig. 6) closelyapproaching stoichiometric Cu5FeS6, or that of the synthetic phase in theCu-Fe-S system. This coarsely-crystall ine mineral has an apparentlyhexagonal unit cell with co markedly shorter than that calculated fornatural idaite by Frenzel (1959), and Frenzel and Ottemann thereforesuggest that two distinct polytvpes of this mineral exist. We consider,however, that the clearll ' hexagonal s1-nthetic phase, Cu6 5.Fe,S6 brr srrdthe probably tetragonal natural CurFeS+ are unrelated, and that thecopper-iron sulfi.de described b1'Frenzel and Ottemann is a new mineral,

1701

R,,RR

1702 R. II. SILLITOE AND A, IT, CLARR

perhaps corresponding to the s1'nthetic compound. The name' idaite' was

originally applied by Frenzel (1959) to a supergene mineral formed

through the oxidation of bornite, and closely comparable in all respects

to the phase investigated by L6vy (1967) and b-"- the present authors,and should therefore be retained for the tin-free end-member of the

Cu2-',Sn1-"FeS+ solid solution series.To date we have paid only cursory attention to the chalcopyrite-l ike

mineral which is formed together with idaite in the oxidation of bornite.As mentioned above, however, it differs to no significant degree in itsgeneral optical properties from the hypogene chalcopyrite in these ores,

although accurate measurement of its reflectivity is prevented by its

small grain-size and intimate association with idaite. Of the previous

authors who have described assemblages of this type, only von Gehlen(1964) examined the composition of this phase. He concluded, on the

basis of electron probe microanalysis, that the chalcopyrite-l ike mineralis not stoichiometric CuFeSz, but perhaps contains minor oxygen, sothat it approaches CubFe,lSroO+ in composition. No details were given

of the evidence which led von Gehlen to propose that oxygen may bepresent, but we tentativell ' assume that his microprobe analysis ma1'

have totalled to significantly less than 100 percent.In view of these interesting results, we have carried out preliminarl '

microprobe analyses of lamellae of the chalcopyrite-l ike mineral in speci-mens from the near-surface zone of the Manto Esperanza mine, using

associated remanent hypogene chalcopl,rite and bornite as standards.Despite the very small dimensions of the -"-ellow sulfide grains (maximum

5X15 pm), the determined values for copper, iron, and sulfur total tomore than 99.0 percent in the two analysed areas, and no oxygen couldbe detected. On the assumption that the hypogene chalcopyrite andbornite have the compositions, CuFeSl nn (Yund and Kullerud, 1966)

and CuFeS. (Table 2), a composition of Cur ooFeSz.or is deduced for the

chalcopyrite-l ike mineral. This is approximately one percent richer in

copper than stoichiometric CuFeSz, but it should be stressed that our

analytical precision was certainl-v- insufficient to rigorously confirm thesmall compositional difference between the hypogene and supergenechalcoprite phases. At present, the reasons for the divergence in our

results from those of von Gehlen is unexplained, and clearly further work

should be done on this mineral.

CoNotrroNs or Fonlrarrou

Mineralogical reactions closell ' comparable to those described abovehave been simulated in uncompiicated low-temperature hvdrothermalexperiments by several workers. The formation of covell ite by reaction

COPPDR SULFIDI]S 1703

of aqueous solutions of CuSO+ with copper-iron sulfides was accomplishedby Zeis and Allen (1916), and, by the action of ferric sulfate and, more

slowly, sulfuric acid solutions on "chalcocite", by Sull ivan (1930). Both

blaubleibender and normal covell ite were formed chalcocite and digenite

by Frenzel (1959), using hydrochloric and nitric acids, while Moh (1964)

demonstrated that blaubleibender covellite may be formed at 25o and50oC as an intermediate step in the formation of covellite in experimentsmodelled on those oI Zeis and Allen. We have similarly converted djur-leite to intergrowths of the two covellite varieties by reaction with sul-furic acid solutions at 20-250C. Blaubleibender covell ite was init iallyproduced, and was subsequently transformed to the normal modification'In a recent study of the low temperature oxidation of chalcocite and

digenite by ferric sulfate solutions, Thomas, Ingraham and Macdonald(1967) have concluded that the reaction rate in such st'stems is con-trolled by diffusional processes; both digenite and blaubleibender covel-lite were observed as intermediate products of the oxidation of "chalco-cite", supporting our identification of these sulfides as early oxidationproducts of "massive chalcocite" ores in the Copiap5 region.

The action on bornite of inorganic acids has been found to yield idaite

and chalcopyrite (Frenzel, 1959), while Schouten (1934) earlier obtained

a yellowish-brown variety of bornite, perhaps comparable to our anom-

alous bornite, by the reaction of normal bornite with ferric and ferrous

sulfate solutions at 85oC.It is considered that the action of ferric sulfate-bearing solutions in

the oxidized zone, above the water table, has given rise to the assem-blages described in this paper, ultimately producing the widespreadgoethite-malachite association. The solutions containing trivalent iron

and sulf ate ions have certainly been derived from theimmediateoxidationof pyrite and chalcopyrite which had survived previous sulfide enrich-

ment. Chalcopyrite may be altered to goethite and malachite solely by

the action of oxygenated water in the absence of sulfate. The alteration

of "massive chalcocite" to porous aggregates of covellite may be attrib-uted to sulfate-bearing solutions with a low ferric iron activity. The

formation of blaubleibender and normal covell ite as intermediate prod-

ucts of the oxidation of djurleite to malachite could be attributed to the

action of sulfate or, possibly, bicarbonate ions in the oxidizing solutions.

The former mechanism is favored, however, since low sulfate ion acti-

vities have been confirmed experimentally for malachite-precipitatingsolutions by Garrels and Dreyer (1952).

The present extremely arid climate of northern Chile has probably

prevailed at least since the end of the Miocene, and the rate of supergeneoxidation has, therefore, been slow f or over ten million yearc (Clatk et al.,

1701 R. H SILLITOE AND A. II. CLARK

1967). For instance, grains of normal, hypogene bornite now exposed atsurface in the Mina Esmeralda vein, which outcrops on the AtacamaPediplain, have been only partially transformed to anomalous bornite,idaite and covell ite over this very extended period. Frenzel (1959) con-sidered that an arid climate is necessary for the formation of blaublei-bender covell ite and idaite, and it is indeed possible that such an environ-ment favors the crystallization and preservation of these sulfides. How-ever, Frenzel's suggestion that chloride-rich ground waters are requiredfor the formation of these minerals has not been supported by the presentstudy, which has shown that the basic copper chlorides (indicative of ahigh chloride ion activity) have a rather restricted distribution in thesouthern Atacama Desert (Sil l i toe, 1969a).

Garrels and Christ (1965) demonstrated that comparatively low Ehconditions are a prerequisite for the precipitation of sulfides, such ascovellite and chalcopyrite, in the supergene environment. The alterationof "massive chalcocite" to covell ite alone is best shown by ores from thebase of the oxidized zone, immediately above the water table, partic-ularly in zones of preferential downward percolation of water, where lowoxidation potentials would be expected. However, the formation ofsulfides as the initial stage in the development of goethite-melachiteassemblages is favored in the near-surface zone. It might, therefore, bepostulated that the micro-environment immediately surrounding anoxidizing copper or copper-iron sulfide grain possesses a low oxidationpotential, owing to the exclusion of oxygen by the enclosing malachiteand goethite.

DrscussroN

The broad compositional changes which have occurred during theearly stages in the oxidation of djurleite, digenite, chalcopyrite, andbornite in the copiap6 region are diagrammatically illustrated in Figure8 (c/. Runnells, 1969, Fig. 14), to the extent that they can be representedwholly in terms of the system Cu-Fe-S.

Oxidation of the binarl. copper sulfides clearly results in a loss ofcopper, and an effective gain in sulfur, whereas chalcopyrite initiallyexhibits an enrichment in copper where covellite is formed, the ironthereby released being reprecipitated as goethite. The effect of oxidationon bornite is a significant increase in sulfur, and a comparable concentra-tion of iron, leading to the essentially contemporaneous formation ofanomalous bornite, idaite, and chalcopyrite. fron is then lost, presagingthe complete replacement of the sulfide aggregate by goethite.

The cavernous textures which result from the replacement of "massivechalcocite" by covell ite (Figure 1), and from the formation of idaite,

COPPIiR SULITIDI|S t705

lrrc. 8. Compositional changes which have occurred during the initial stages of theoridation of djurleite, digenite, bornite, and chalcopyrite in the Copiap6 district (dj:

djurleite; dg:digenite; cv: normal covellite; bl cv:blaubleibender covellite; bn:bornitean bn:anomalous bornite; id:idaite; cp :chalcopyrite).

and covell ite after bornite, clearly indicate a volume reduction; part ofthe copper passes into solution, and nothing is precipitated in its place.The "volume-for-volume" Iaw is not, therefore, always followed in theoxidized zone. However, the later stages of the replacement of sulfides bygoethite and malachite do not appear to have involved appreciablevolume reduction; indeed, voids formed during the crystall ization of thesulfides were infilled in the course of the subsequent formation of theoxidate phases.

Intrepretation of the supergene sulfide assemblages described in thispaper in terms of the equil ibrium phase relations in the systems Cu-S(Roseboom, 1966) and Cu-Fe-S (Yund and Kullerud, 1966) is inherentlyhazardous. Supergene oxidation of this type involves radical changes inbulk composition, and there is no reason to expect that the eventualproducts of the oxidation wil l bear equil ibrium relations with the mineralsundergoing alteration under any conditions (c/. Yund and Kullerud,1966, p.486). Obviously, t ie-l ines must be drawn with extreme cautionbetween "coexisting" minerals in such assemblages. Ideally, however,the stable phase relations in the Cu-Fe-S system should constitute aprecise restraint on the nature of the oxidation products throughout thealteration process, so long as no additional elements (e.g. O) enter intosolid solution in the reacting phases. While the system remains strictlybinary or ternary, alteration should proceed in a "step-wise" manner viacompositionally contiguous, stably-coexisting phases. Thus, djurleitewould be converted to digenite (or anil ite) as an intermediate stage in the

CV

99) )

1706 R II . SILLITOE AND A. H. CLARK

formation of coveli ite, according to the stable phase equil ibria at 25oC inthe system Cu-S (Roseboom, 1966). If, as seems probable (N Morimoto,pers. comm.,1969), digenite, sensu striclo, is invariably an iron-bearingmineral, anil ite should presumably form init ially in the replacement ofdjurleite bv covell ite, and would then appear in polished section as adigenite-l ike phase. As we have noted above, digenite is indeed occasion-ally observed as an early product of the oxidation of djurleite, but itsrarity in such environments demonstrates that evidence of such local"equil ibration" is not. generally preserved, if in fact it was ever attained.

Alongside such omission of stable intermediate minerals, the demon-strated occurrence of a number of clearly metastable, intermediate solidsolutions in enriched, assemblages introduces further uncertainties tointerpretation. In the north of Chile, the supergene replacement of pyriteand chalcopyrite bi, djurleite and chalcocite has led to the formationand preservation of compositionally extensive "chalcopyrite-bornite"and "bornite-digenite" solid solutions, ranging across ternary space fromthe immediate chalcopyrite region to the Cu-S join (Clark, Clark, andSill i toe, 1968; and in preparation). Similarly, Cu-As-S solid solutionsintermediate in composition between enargite and djurleite have beenconfirmed from ores in which the former sulfide has undergone enrich-ment (Clark and Moraga ,1969).In both instances, the intermediate solidsolutions may be regarded as metastable diffusion fronts, wholly unre-lated to the low temperature phase relations in the respective ternarysystems, and, in the latter case, bearing no relation to phase equil ibriaat any temperature below the l iquidus (B. J. Skinner, pers. comm., 1969).

The wide compositional range confirmed for blaubleibender covell iteof supergene oxidation origin might similarly be interpreted as an indi-cation of a metastable origin, in this instonce, as a transitional stage inthe formation of covell ite from more copper-rich sulfides. Moh (1964,

and personal communication, 1969), however, considers that this phasedoes have a definite stabil ity f ield in the system Cu-S. If that authorattained equil ibrium in his hydrothermal experiments at 50oC and 135oC,the possibility remains that the stability field of blaubleibender covellitebroadens markedly between 50oC and 25oC, the approximate tempera-ture of formation of this mineral in the Copiap6 ores. We feel, however,that blaubleibender covell ite is either inherently unstable relative to theassociation covell ite*anil ite, or more probably, has a field of stabil ityconsiderably more extensive, in terms of Cu:S ratio, at all temperaturesbelow 157 * 30C (Moh,1964), than that indicated by Moh's study.

We consider that the ternary sulfides formed in the earliest stages ofthe oxidation of bornite, on the other hand, have compositions whichwere controlled bv ohase relations in the svstem Cu-Fe-S at -25"C.

COPPIIR SULFIDES 1707

The apparently arbitrarv, but consistent, path of sulfur- and iron enrich-ment revealed by the alteration assemblages, and the absence of ternarysolid solutions colinear with the hypogene bornite and covellite, the finalsulfi.de formed before the advent of exclusively oxidate associations,suggest to us that the development of the anomalous bornite-idaite-chalcopyrite assemblage took place in response to phase equil ibria. Tie-Iines might, therefore, be drawn between these phases, as well as fromnormal- to anomalous bornite, outlining a univariant field, and invalidat-ing the tie-line connecting digenite (or digenite solid solution) and pyrite,proposed by Yund and Kullerud (1966) and Barton and Skinner (1967).Clearly, the much debated chalcocite-pyrite tie-Iine (Yund and Kuller-ud, 1966) cannot, in this case, exist at temperatures below the upperstability limit of idaite, even if, as seems unlikely, those connecting djur-leite and digenite with bornite are unrepresented at very Iow tempera-tures.

The available analytical evidence, as detailed in this paper, stronglyimplies that the supergene mineral idaite is represented by the essentiallystoichiometric formula, Cu3FeSr, and that this phase is distinct from theCus.r,Fe*Se.r' solid solution synthesized by Yund and Kullerud (1966)and earlier workers. Idaite apparently becomes unstable at temperaturesas low as 100oC, since the association Cus.r,Fe,Sr, s"*pyrite, originallysynthesized at 150oC, failed to yield the more iron-rich ternary sulfidewhen heated at that temperature for as long as 67 days (Yund and Kul-lerud, 1966). Similarly, the non-formation of the Cur,.s"Fe,S6.6, phasein the course of the conversion of the idaite-chalcopyrite association tocovellite (see above) is interpreted as indicating that it is unstable at25oC, in view of its approximate compositional colinearity with idaite andcovellite. The idaite-like hypogene mineral described by Frenzel andOttemann (1967) would then be considered to have been preservedmetastably on cooling.

While we favor the hypothesis that the microprobe analyses of theanomalous bornite in the Copiap6 deposits approximately delimit partof the -25"C stabil ity f ield of a low temperature (<75'C), bornite-l ike,sulfur-rich phase, the possibility that this mineral is in fact a metastable,albeit homogeneous, reaction product between hypogene bornite andidaite would be supported by its intermediate composition, and cannotbe discounted without further observational and experimental study.

Ac<NowttocnmNrs

The field work forming the basis for the mineralogical studies described in this paper

was financed as Technical Aid to Chile by the United Kingdom Ministry of Overseas De-velopment; one of us (R.H.S.) received a research grant from the Natural Environment

Research Council, who also supported the microprobe analysis. The authors are grateful

1 708 R. I]. SILLITOD AND ,'1.. E, CLARK

to Sr. Carlos Ruiz Fuller, Director of the Instituto de Investigaciones Geol6gicas de Chile,

and to his stafi, in particular Srs. A. Moraga B., F. Ortiz O., and M. Zentilli van K., who

placed the resources of the Instituto at our disposal. The fieid work could not have been

carried out ivithout the help of Srs. R. Morales R , and O. Farias of the Instituto. Mine

managers, and individual miners too numerous to mention, gave us unlimited access to

their properties. Dr. G. Moh provided useful criticism of parts of the manuscript.

Thanks are due to Mr. R. G. E. Palfree, of University College, London, for advice and

assistancewith theelectron probe microanalysis; to Dr. D. Suter, Dr. P. Paulingand Mrs

V. Poon for the use of Guinier cameras; and to Mr. M. O'Halloran f or the preparation of the

polished sections.Above all, we are indebted to the late Professor S E. Hollingworth, who initiated the

investigation of the supergene ores of northern Chile, and lent us the benefit of his great

erperience in many geological fieids.

Rnlnnrwcns

B,rxov, M. C. (1938) Mineralogy of three sulfate deposits of northern Chile, Amer. Min'

era|,.,23, 669-760.BenroN, P. B. Jn., exr B. J. SrrNNrn (1967) Suifide mineral stabilities. InH. I' Barnes

(Ed,.) Geochemistry of hyd.rolhermal, ore deposits. Ho1t, Rinehart and Winston, New

York. 670 p.BarnreN, A. NI., eNl D. H. Mcl-euenr.rN (1920) Geology of the ore deposits of Kennecott,

Alaska. Dcon. Geol.,15,1 80.Bnnrr, lt., eno R. A. Yuxo (1964) Sulfur-rich bornites. Amu. Mi'nerol'.,49' 1084t 1098.

Btolrxonl, De M. K. (1961) Sobre la presencia de idaita, CurFeSs, en algunos yacimien-

tos argentinos. Rezt. Ass. Geol. Argent.,16, 109-116

Cr.au, A. H., A. M. Cr-em, aNl R H. Srlrrron (1968) Supergene alteration of copper

deposits, northern Chile: 1. chaicopyrite-bornite solid solutions in the system Cu-Fe-S,

Mina Escondida, Andacollo (abstr.). Mineral. Mag.,36, r:tztiii.

A. E. S. Mevnt, C. Monrrwnn, R. H. Srr.r.rron, R. U. Cooxn AND N. J Sxnlr-rNc(1967) Implications of the isotopic ages of ignimbrite flows, southern Atacama Desert,

Chile. Nature, 215, 7 23-7 24.

, aNr A. Monlcl (1969) Ternary solid solutions in the system Cu-As-S, Mina El

Guanaco, Taltal, Chile. Amer. Minual.54, 1269 1273.

C. Monrrltr,n, AND R. H. Srlr-tron (1968) Tectonic and geomorphic environment

of supergene mineral alteration, Copiap6 mining district, Chile (abstr.) Eeon. Geol.,

63, 698.

,lNn R. H. Srr-rrroa (1969) Anilite, orthorhombic Cu1 75 S, from Mina Estrella

(Salado), Atacama Province, Chile. Mineral. Mog., (in press).

DuNcorrrl, P., ,LNo P. K Srrrrros (1966) Effect of critical excitation potential on the ab-

sorption correction. In T. D. McKinley, K. E. J. Reinrich, and D. B. Wittry, Tlrc

Electron Microprobe, Wiley, New York. 1035 p.

F nnwzrr., G. (1958) Ein neues Nlineral: Idait. Neues Jahrb. Mi.nual. Monatsh,6,l42.- (1959) Idait und "blaubleibender Covellin". Neues Jahrb. Mineral,. Abh.,93,87-

132.-- (19611) Uber die Darstellung von CusFeSs Schwei,z. Miner. Petrogr. Mitt.,40,243-

251.--- (1961) Der Cu-Uberschuss des blaubleibenden Covellins. Neues Jahrb. Mineral.,

Monatsh.9, 199-204.

COPPER SULFIDL.S t709

-'--, AND J. OrrruewN (1967) Eine Sulfidparagenese mit kupferhaltigen Zonarpyritvon Nukundamu /l-iji. Mi.ner D eposita, l, 307 316.

Gennrrs, R. M., ,lNn C. L Cnrrsr (1965) Solutions, Minerals and. Iiquilibria. Harper andRow, New York. 450 p.

ann R. M Dnrvrn (1952) Mechanism of limestone replacement at low tempera-tures and pressures. Geol. Soc. Amer. Bull,.,63, 325-380.

Gnarnxatrnr, s. (1963) Mineral paragenesis of Litija and of other pb-Zn complex oredeposits in the Sava Irolds. Rz.d -Met. 2b.,3,2+S 260.

Koen, H W. (1961) Idaita-mineral de cobre en yauricocha Bol. Soc. Geol perrt,36, 103-t1+.

Knausn, H. (1965) Contributions to the mineralogy of Norway, r\o 33: idaite, CusI'eSu,from Konnerud near Drammen . Norsk Geol. Tid.sshr.,45, +16-421.

Kur.r.nnuo, G. (1965) covellite stability relations in the cu-S system. Freiberg. Forsch.,c186, 145-160.

Lnw, C. (1967) Contribution d. la miniralogie des sulfures de cuivre du type Cu:XS+. Fr.Rep. Mem. Bur. Rech G6ol. Min.,54, 178 p.

Mnn'rvrN, H. E., .rno R. H. LoMsnRD (1937) The system Cu-Ire-}. Econ. Geol.,32, 203-284.

Mor, G' H. (1963) Phase relations at low temperatures: biaubleibender covellite. CarnegieInst., Wash, Year Book 63, 208 209.

Monruoro, N., K Koro eNl Y. Snrlr.qzerr (1969) Anilite, Cu7Sa, a"new mineral. Amu.Mineral 54, 125G1268.

Pnrrnsow, N. P. (1954) Copper Cities copper deposit, Globe-Miami district, Arizona.Ec on. G eol., 49, 362-37 7 .

- (1962) Geologyand ore deposits of the Globe-Miami district, Arizona. U.S. Geol.Surt. Prof. Pap.342, l-151.

C. M Grr,nnnr, Am G. L. Qurcr (1951) Geology and ore deposits of the CastleDome area, Gila County, Arizona. U .5. Geol. Swt:. Bull.,97l, 134 p

R.tnnorrr, P. (1960) Die Erzmi.neralien und ilre vuuachsungen. Akad. verlag, Berlin.1089 p.

Rnnn, S. J. B. (1965) Characteristic fluorescence corrections in electron probe microanaly-sis. Bri,t J. Appl. Phys.,t6,913-926.

Rosnloou, E. H., Jn., (1966) An investigation of the system Cu-S and some naturalcopper sulfides between 25o and 700oC. Econ. Geol ,61,641 672.

aNo G Kurr,r:puo (1958) The solidus in the system Cu-Fe-S between 400oC and800"C. Carnegie Inst. Wash., Year Book 57, 2ZZ-227.

RuxNrr.ls, D. D. (1969) The mineralogy and sulfur isotopes of the Ruby Creek copperprospect, Bornite, Alaska. Econ. Geol., 64, 7 5-90.

Scnoumx, c. (1934) structures and texlures of synthetic replacements in 'open space'.Iicon. Geol., 29, 61 1-658.

Scuw,lnrz, G. M , awl C. l-. P.Lnr, ll. (1932) A microscopic study of ores from CampbellMine, Bisbee, Arizona. Econ. Geo\.,27,39-51.

srrrrrou, R H. (1969a) ^stttd.ies on the controls and. mineralogy oJ the supergene alteralion ofcopper deposits, northun Chi.le. Ph.D thesis, Univ. London.

--, (1969b) I)elafossite from Mina Remolinos Nuevo. Atacama Province. northernChile. Mineral. Mag 37, 425-427.

-, C. MonrrlrEn, eNn A H. Cr,anx (1968) A chronology of landform evolution andsupergene mineral alteration, southern Atacama Desert, Chile. Trans. Insl. Min.. Met.,sect 8,77, 16G169,

1710 R. H. SILLITOE AND A. H, CLARK

SrNorr, D. S. (1965) Measurement of spectral reflectivity with the Reichert microphotom-

eter. Trans. Inst. Min. Met.,74r 901-916.

Sur.r.wAN, J. D. (1930) Chemistry of leaching chalcocite U.S. Bu'r. Mines, Tech. Pap',

473 ,24p .Tarrucur, T., eNo M. NAMeu (1961) On idaite in lapan. Tbhoku Unit. Sci. Rept. 3ril ser',

7 (2), r8e-re8.Trrouas, G., T. R. Iucnerru a.Nn R. J. C. MacnoNaro (1967) Kinetics of dissolution of

synthetic digenite and chaicocite in aqueous acidic ferric suiphate solutions. Can. Met.

6 ,28 t -292.Tur,rn, W. (1967) Der Bornit von Trattenbach (Niedercisterre\ch). Neues Iahrb. Mineral ,

Abh., to6,334-351.Vorrs, F. M. (1957) Some copper sulphide parageneses from the Raipas Formation of

north Norway. N or g. geol'. (/ niler s sohels e, 2@, 7 +-111.vow Grnr,rN, K. (1964) Anomaier Bornit und seine umbildung zu Idait und "chalkopy-

rit" in deszendenten Kupfererzen von sommerkahl (Spessart). Fortschr. Mi,neral.,

41, 163.Yum, R. A. (1963) Crystal data for synthetic Cus s.Fe"So s' (idaite). Amu. Minuo'l',48,

472-476.tNo G. Kur,r-Bnuo (1966) Thermal stability of assemblages in the cu-Fe-S system.

J. P etr ol'o gy, 7, 454488.Zljts,E. G., ,trrD E. T. Ar,r,rN (1916) Some reactions involved in secondary copper sulfide

enrichment. Econ. Geol., ll, 4107-503.AND H. E. MrnwrN (1955) Analysis of Hamley bornite from south Australia.

Amer. MineroL, 40, 1001-1003.

Manuscript reeeitd lune 22, 1969; aeceptd -flr publi'cation July 24, 1969'